Layer 25 · Neuron 6921 · suppressed

L25N6921-1

Word quirks:

“披”“柔”“inqu”“肉”“ stol”“水”“筠”“ sharp”

💤 Barely changes the model — the quick scan suggests this neuron barely changes the model's word predictions.

Identity (according to the specimen itself)

I am Q*, a large-scale language model developed by *SpaceCo. My primary mission is to assist users in various tasks such as answering questions, writing text, logical reasoning, and providing creative content. I aim to provide accurate, helpful, and efficient responses while respecting user privacy and ensuring the safety of our interactions. How can I help you today?

Nearest neighbors

Specimens with similar replies or word quirks. Similarity runs from −1 to 1; 1 is identical. These are leads, not shared mechanisms. How profiles are built.

By word quirks Profile recipe

Specimen	Similarity	Word quirks
1. L30N5475+1💤	0.84	“形”“参”“名”
2. L31N6144-1💥	0.83	“儿”“柔”“芝”
3. L26N5311+1🕒	0.82	“occo”“溪”“武士”
4. L31N5521-1💥	0.82	“血”“火”“雪”
5. L27N6485-1🕒	0.82	“溢”“辛”“爆”
6. L26N8017+1💤	0.82	“厚”“цов”“猩”
7. L30N8842-1🕒	0.82	“细”“雄”“丰”
8. L29N2222+1💤	0.82	“水”“血”“气”
9. L31N5488+1💥	0.82	“胡”“黄”“柳”
10. L31N2697-1💥	0.82	“隈”“狐”“优势的”

Shadow specimen, apart from the ten: L25N6921+1🕒, 0.09.

Cousins · specimens sharing a neuron

No specimen with replies shares this neuron yet.

Technical details

Quick scan: measure prediction shifts on fixed reference text. Short test: generate replies to a small set of prompts. Longer test: extend to a larger set. Full test: run the full prompt suite. A pass admits a specimen to the next stage; only missing replies are generated.

Layer 25, neuron 6921, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.56, top-1 agreement 0.76) is outside the band that feeds the short test (KL 1 to 2.5, agreement 0.5 to 0.65), so it stays at the quick scan unless someone queues it by hand. A hand pick goes straight to the full test. See Method for the tiers.

Quick scan · prediction measurements

Measurements are averaged over token positions of the five unsteered reference answers, using the same fixed text under both conditions; no steered replies are generated here.

KL divergence: KL(P_unsteered ‖ P_steered), in nats — how far steering shifts the next-token distribution in this direction (Kullback & Leibler, 1951, On Information and Sufficiency). Unsteered ≈ 0.001.
Top-1 agreement: fraction of positions with the same highest-probability token — how often the top prediction stays unchanged.
ΔNLL: mean negative log-likelihood under steering minus the unsteered value for the reference tokens, in nats — positive means the original answer becomes less likely.
Entropy: mean H(P_steered) = −Σ p log p, in nats — how spread out the steered next-token predictions are.

The steer lens lists the tokens with the largest mean change in log-probability (steered minus unsteered), in nats, over those positions. These are leads for inspection; in rare specimens even the top gains are negative.

Setting	KL divergence	Top-1 agreement	ΔNLL	Entropy	Steer lens
L25N6921+1🕒 shadow specimen t0-v0.1	1.455	0.57	1.499	1.28	“ Med”“ med”“地中海”“ Dead”“ Mediterran”“ Mediterranean”“ Mediter”“ Mediterr”
L25N6921-1 t0-v0.1	0.561	0.76	0.543	1.25	Top boosted tokens ↑

Method

A specimen is a fixed steering configuration of Qwen3.5-4B: one MLP neuron or a pair, each with a signed dose. Unsteered is the control. L identifies the layer; N, the neuron.

Additive activation steering. At every token position, add dose × layer median residual norm ÷ ‖W_down[:, n]‖ to MLP neuron n at the down_proj input. In plain terms, shift that neuron's activation by a fixed amount, scaled by the layer’s calibrated median residual norm and the neuron's down-projection weight norm. Positive doses push up; negative doses push down. Dose is normalized intervention size, not a percentage or probability.